Download A prospective pilot study of curative-intent stereotactic body radiation

650
A Prospective Pilot Study of Curative-intent
Stereotactic Body Radiation Therapy in Patients
With 5 or Fewer Oligometastatic Lesions
Michael T. Milano, MD, PhD1
Alan W. Katz, MD, MPH1
Ann G. Muhs, BS1
Abraham Philip, CMD1
Daniel J. Buchholz, MD2
Michael C. Schell, PhD1
Paul Okunieff, MD1
BACKGROUND. It is hypothesized that oligometastatic disease represents a state of
1
RESULTS. The 2-year overall survival (OS), progression-free survival (PFS), local
potentially curable, limited metastases. Stereotactic body radiation therapy
(SBRT) is an option for patients who are not amenable to or do not want resection.
METHODS. From 2001 to 2006, 121 patients with 5 detectable metastases were
enrolled in 2 prospective studies that used curative-intent SBRT. Most patients
were treated with 10 fractions of 5 Gray. Stereotactic radiosurgery was offered to
patients with brain metastases.
Department of Radiation Oncology, University of
Rochester Medical Center, Rochester, New York.
control (LC), and distant control (DC) rates were 50%, 26%, 67%, and 34%,
2
respectively; and the respective 4-year rates values were 28%, 20%, 60%, and
Department of Radiation Oncology, M. D.
Anderson Cancer Center, Orlando, Florida.
25%. A greater net tumor volume predicted significantly worse OS, PFS, LC, and
DC. Patients with breast cancer fared significantly better with respect to OS, PFS,
LC, and DC; and patients with adrenal metastases had significantly worse OS,
PFS, and DC despite the small number of such patients enrolled. Neither the
number of metastatic lesions nor the number of organs involved was a significant predictor of outcome. Among 45 patients who remained alive at the last
follow-up, 29 patients had no evidence of disease, including 23 patients with
2 years of follow-up.
CONCLUSIONS. Oligometastatic disease is a potentially curable state of distant
cancer spread. In this hypothesis-generating analysis, patients with less volume
burden of their metastatic disease and those with primary breast cancer fared
better. SBRT delivered with curative intent in patients with limited metastases
should be investigated further. The Southwest Oncology Group is developing a
prospective protocol to treat women who have limited breast cancer metastases
Paul Okunieff received grant support from BrainLAB AG (Heimstetten, Germany) between April 1,
2000, and March 31, 2005.
We thank the family of Nathaniel Rutter for their
support. We also thank Laura Brumbaugh for
editorial assistance.
Novalis, ExacTrac, and BrainSCAN are trademarks
of BrainLAB, AG, Heimstetten, Germany.
Address for reprints: Michael T. Milano, MD, PhD,
Department of Radiation Oncology, University of
Rochester Medical Center, 601 Elmwood Avenue,
Box 647, Rochester, NY 14642; Fax: (585) 2751531; E-mail: [email protected]
Received June 19, 2007; revision received
August 21, 2007; accepted August 22, 2007.
ª 2007 American Cancer Society
with SBRT. Cancer 2008;112:650–8. 2007 American Cancer Society.
KEYWORDS: disease burden, local control, oligometastases, stereotactic body
radiation.
T
he clinical state of oligometastatic disease was proposed in 1995
by Hellman and Weichselbaum.1 They hypothesized that, in
some patients with a limited number of clinically detectable metastatic tumors, the extent of disease exists in a transitional state
between localized and widespread systemic disease. In this model,
oligometastatic disease has the potential of progressing to widespread metastatic disease. Thus, local control (LC) of oligometastases may yield improved systemic control.1,2 An alternative
hypothesis is that oligometastatic disease represents a clinical manifestation of few detectable lesions in the setting of widespread
occult disease. Such a model argues that local therapy alone
DOI 10.1002/cncr.23209
Published online 10 December 2007 in Wiley InterScience (www.interscience.wiley.com).
Curative SBRT for Oligometastases/Milano et al.
probably would not be curative. Arguably, both models may be correct, and patients may exist in a spectrum between orderly metastatic progression and
widespread occult disease, a model that also has
been proposed for primary cancer spread.1,3
It has been proposed that progressive growth of
oligometastatic tumors beyond a threshold size
results in an exponential rise in the risk of further
metastatic progression.4 In the absence of micrometastatic disease or in the situation in which micrometastatic disease is controlled by systemic therapy,
aggressive treatment of oligometastatic disease can
be considered curative-intent therapy, as evidenced
by the prolonged disease-free survival observed in
patients with resected oligometastatic tumors.1,5–10
Radiotherapy is another means with which to control
oligometastatic tumors, particularly in patients who
cannot tolerate or do not want surgery or when
tumors are situated in areas in which resection
would result in unacceptable morbidity. Several institutions use hypofractionated stereotactic body radiotherapy (SBRT) to treat oligometastases.11
SBRT implies the use of a 3-dimensional frame
of reference to localize the tumor accurately. Because
setup uncertainty and tumor movement are reduced
significantly, the planning target volume (PTV) margins can be reduced compared with the PTV margins
of standard conformal radiation.12 Consequently,
SBRT limits the radiation dose to normal tissues and
allows delivery of higher doses per fraction. SBRT is
well suited for oligometastatic disease, because
aggressive fractionation can be used to attempt
improved disease control with acceptable toxicity.11,13
The American Association of Physicists in Medicine
(AAPM) Task Group 101 report will review the available SBRT commercial products and relevant literature on SBRT, outline guidelines for commissioning
SBRT, and develop standard protocols for quality
assurance and SBRT treatment planning and delivery.14 A systematic review and guidelines for intracranial stereotactic radiosurgery (SRS) are offered in the
AAPM Task Group 42 report.15
Since 2001, the University of Rochester has used
hypofractionated SBRT to treat patients with oligometastatic disease. The outcomes in patients who
were treated for liver metastases16 and lung metastases17 have been published recently. In an earlier
analysis of patients with metastatic prostate cancer,
our group demonstrated that patients with 5 metastatic lesions fared significantly better than those
with >5 lesions.18 Thus, extrapolating from these
data, our protocol was designed so that the treatment of patients with >5 lesions was considered palliative. The focus of the current study was patients
651
who were treated with curative-intent SBRT for 5
metastatic lesions, and the objective of the study was
to determine patient and tumor variables that predict
a better outcome.
MATERIALS AND METHODS
Between February 2001 and December 2006, 160
patients were enrolled on 2 University of Rochester
Cancer Center (URCC) pilot studies using SBRT to
treat limited oligometastatic disease and to assess
the feasibility and outcome of such an approach.
URCC U8700 was designed specifically to treat
patients with metastatic breast cancer without brain
metastases, whereas URCC U9700 included patients
with any primary cancer and any involved sites.
Patients on URCC U9700 with limited brain metastases were offered SRS. Both studies were approved by
the University of Rochester Research Subjects Review
Board, and all patients signed informed consent.
Prior treatment of metastatic tumors (including
radiation or surgery) did not exclude patients from
the study unless the treating physician determined
that radiation could not be delivered safely. Prior
chemotherapy was allowed; in fact, most patients
had received 1 courses of chemotherapy before
they enrolled. Eligibility requirements included age
18 years and a Karnofsky performance status 70.
All patients were assessed with a diagnostic body
computed tomography (CT) scan to identify and discern metastatic lesions. For brain lesions, magnetic
resonance imaging (MRI) was required for diagnostic
and planning purposes. MRI was also used to assess
liver metastases and bone metastases in some
patients. Many patients underwent positron emission
tomography (PET) scanning. Patients with bone metastases generally had a bone scan and/or PET scan
in addition to CT scans.
No more than 5 lesions were treated in a single
course of treatment. The treatment of 5 lesions in
patients who presented with >5 lesions was considered palliative. Patients who presented initially
with >5 clinically apparent, oligometastatic tumors
may have received treatment to all of their lesions
over 2 courses of treatment, but these patients
were excluded from the current analysis, because
they were treated with palliative intent.
The current study includes patients who were
enrolled in either URCC U8700 or U9700 with 5
metastatic lesions at the time of enrollment who
were treated with curative intent. In total, 121 of 160
enrolled patients are included in the current analysis.
Patients who experienced local failure after SBRT in
1 sites or who developed further metastatic disease
were allowed to receive additional courses of SBRT.
652
CANCER
February 1, 2008 / Volume 112 / Number 3
SBRT Technique
The technique of SBRT employed at the University of
Rochester is discussed in detail in previous publications16,17,19 and is summarized briefly here. During
the initial simulation and with all treatments, patients
were immobilized with a vacuum bag, and treatment
setup was reproduced by using a relaxed, end-expiratory breath-hold technique and the Novalis ExacTrac
patient-positioning platform (BrainLAB, AG, Heimstetten, Germany). Treatment planning was performed by using the BrainSCAN system (BrainLAB).
The gross target volume (GTV) was contoured on axial
CT images; MRI and PET scans, when available, were
fused with the planning CT scan for more accurate
delineation of the GTV. No margin was added for the
clinical target volume. The PTV was generated with a
minimum GTV expansion of 10 mm in the craniocaudal direction and 7 mm in other directions. We previously demonstrated that these margins allow for
coverage of 2 to 3 standard deviations of motion.19
Treatment was prescribed to the 100% isodose
line, with the 80% isodose line covering the PTV.
SBRT was delivered by using conformal arcs or multiple, fixed, coplanar beams. The dose per fraction and
total dose were determined by using the dose volume
histogram of organs at risk with a preferred schedule
of 50 Gray (Gy) in 5-Gy fractions over 2 weeks. The
normal tissue constraints for the liver and lung are
detailed in our previous publications.16,17 Concurrent
chemotherapy with a nonanthracycline-containing
regimen was permitted.
SRS Technique
Immobilization was achieved with a BrainLAB head
frame attached to the patient’s skull under local
anesthesia. SRS was delivered with the Novalis linear
accelerator, which was equipped with micromultileaf
collimators, using 6-megavolt photons. A CT scan
was obtained in the axial plane with the Novalis
localizer frame in position. The GTV was defined
based on MRI and CT images. A margin from 0 mm
to 1 mm was added to the GTV to create the PTV.
The prescribed isocenter dose was from 10 Gy to 20
Gy with the 80% isodose line covering the PTV.
BrainLAB planning software was used to generate the
treatment plan, which employed multiple arcs with
fixed-shaped fields. The dose to the critical structures
was evaluated carefully before finalizing the plan.
Quality assurance was carried out as discussed in a
prior publication.20
Follow-up
Follow-up visits were planned 1 month after completing SBRT and every 3 months subsequently for
2 years. Thereafter, intervals ranged from 3 months
to 6 months, based on physician preference. Patients
underwent diagnostic imaging studies before all follow-up visits after the initial 1-month visit. Toxicity
in patients was evaluated by using the Common Terminology Criteria for Adverse Events criteria (version
3.0).
Endpoints
Several time points were recorded for each patient,
including the dates of primary cancer diagnosis, metastatic disease diagnosis, and first appearance of
lesions treated on protocol. Overall survival (OS) and
progression-free survival (PFS) were calculated by
using Kaplan-Meier actuarial survival analyses, with
survival and failure times defined from the day of
enrollment until either an event or last follow-up.
Local failure either was scored as an event if any
treated lesion grew by 20%, based on the Response
Evaluation Criteria In Solid Tumors criteria21 or was
confirmed pathologically. Distant failure was scored
as an event if a patient progressed with distant metastases beyond what could be treated with curativeintent (ie, 5 metastases or incurable disease). Progression was defined as local or distant failure. Significant (P < .05) variables on univariate analyses
(UVA) were tested with multivariate analyses (MVA).
MVA were performed by using a Cox proportional
hazards regression model. Stata software was used
for all data analysis (version 9.2).
The net GTV is the sum of GTVs based on the
contoured volumes on the planning CT scan.
Because the net GTV was assessed at the time of
enrollment, previously resected metastases were not
included in the net tumor volume; likewise, changes
in the tumor volume resulting from prior systemic
therapy were not accounted for. Tumor grade was
not included in these analyses because of the different spectrum of primary cancers involved.
RESULTS
Patient Characteristics
Patient characteristics are summarized in Table 1.
One patient had brain-only disease. Nine patients
had bone-only oligometastatic disease, and 8 of
those patients had primary breast cancer. The
primary site was controlled at the time of enrollment
in all patients, with the following exceptions: Four
patients with primary nonsmall cell lung cancer presented initially with multiple lesions (including thoracic lymph nodes in 2 patients); 3 patients with
primary T1 nonsmall cell lung cancer presented initially with synchronous, metastatic disease (adrenal,
Curative SBRT for Oligometastases/Milano et al.
TABLE 1
Patient Characteristics at Initial Presentation of
Oligometastatic Disease
Characteristic
No. of patients (%)
No. of patients
Age, y
Median [range]
Mean SD
Enrolling institution
University of Rochester, Rochester, NY
M. D. Anderson Cancer Center, Orlando, Fla
Primary cancer
Breast
Colorectal
Lung, head and neck, or esophagus
Pancreas, biliary tract, or liver
Sarcoma
Other
Primary histology
Adenocarcinoma
Squamous cell carcinoma
Sarcoma
Hepatocellular carcinoma
Renal cell carcinoma
Carcinoid
Other
Previously had >5 metastatic lesions
Prior curative-intent local therapy
Resection
Radiofrequency ablation
Radiation therapy
Cranial stereotactic radiosurgery
Sites involved with oligometastatic disease
Lung
Thoracic lymph nodes
Liver
Pelvic or abdominal lymph nodes
Brain
Adrenal glands
Bone
No. of oligometastatic lesions
1
2
3
4
5
No. of involved organs
1
2
3
Sum of GTVs, mL
Median [range]
Mean SD
121 (100)
60 [34–88]
58 12
111 (92)
10 (8)
39 (32)
31 (26)
23 (19)
7 (6)
7 (6)
14 (12)
89 (74)
7 (6)
7 (6)y
3 (2)
3 (2)
3 (2)
9 (7){
21 (17)
36 (30)
30
4
3
3
50 (41)
24 (20)
54 (45)
4 (3)
5 (4)
2 (2)
15 (12)
37 (31)
32 (26)
28 (23)
12 (10)
12 (10)
92 (76)
25 (21)
4 (3)
28 [0.3–422]
52 75
SD indicates standard deviation; GTV, gross tumor volume.
Other primary cancers included carcinoid (n 5 3), urinary bladder (n 5 3), renal (n 5 3), adrenocortical, ovarian, endometrial, endocervical, and melanoma.
y
Sarcoma subtypes included leiomyosarcoma (n 5 3) and high-grade undifferentiated, synovial cell,
spindle cell, and Ewing sarcoma.
{
Other histologies included transitional cell carcinoma (n 5 2), poorly differentiated nonsmall cell
lung cancer (n 5 2), large cell neuroendocrine carcinoma, small cell carcinoma, cystadenocarcinoma, and cortical carcinoma.
653
bone, and brain); and 2 patients presented with multiple liver lesions (carcinoid and hepatocellular carcinoma). The primary site(s) was treated in each of
these patients with SBRT.
Thirty-five patients presented with metastatic
disease during their initial diagnostic workup for
cancer. Of these 35 patients, 27 presented with the
lesions that were treated on the current protocol. For
the remaining 8 patients, the interval between the
initial diagnosis of metastatic disease to the appearance of the metastases treated on protocol was
from 5 months to 91 months (median, 22 months).
During the interval, all 8 patients received chemotherapy, and all 8 patients underwent at least 1
metastasectomy.
Eighty-six patients developed metastatic disease
from 3 months to 136 months (median, 27 months)
after their initial diagnosis of cancer. In 65 of these
86 patients, the lesions that were treated on the current protocol were evident at the time of proven metastases. For the remaining 21 patients, the interval
between the initial diagnosis of metastatic disease
and the development of the metastases treated on
protocol was from 3 months to 56 months (median,
20 months). In this interval, all but 1 patient received
curative-intent therapy for oligometastatic disease,
including metastasectomy (n 5 16 patients), radiofrequency ablation (n 5 2 patients), fractionated
radiation (n 5 2 patients), and cranial stereotactic
radiosurgery (n 5 1 patient). One of these 21
patients was treated with whole-brain radiation
for >5 brain metastases, which no longer were
evident at the time of enrollment.
The time from the first appearance of the oligometastatic lesions treated on protocol to enrollment
ranged from <1 month to 96 months (median, 7
months). Ninety-eight of the 121 patients received
systemic therapy for metastatic disease before enrollment.
Thirty-six patients underwent curative-intent
treatment for oligometastatic disease at some point
before enrollment (see Table 1). Twenty-one patients
were diagnosed with 5 metastases at some point
before enrollment. These patients received systemic
therapy and/or radiation, which ultimately resulted
in 5 detectable metastases at the time of enrollment.
Toxicity of SBRT
No patient experienced grade 4 or 5 toxicity, and
only 1 patient experienced grade 3 toxicity (described
below). No patient experienced grade 2 fatigue or
desquamation. The toxicity in patients who received
SBRT for lung and mediastinal tumors was addressed
654
CANCER
February 1, 2008 / Volume 112 / Number 3
FIGURE 1. Kaplan-Meier actuarial overall survival and progression-free
FIGURE 2. Kaplan-Meier actuarial local control and distant control.
survival.
previously.17 The only grade 3 toxicity observed was
a nonmalignant pleural and pericardial effusion. The
toxicity in patients who received SBRT for liver metastases also was discussed previously.16 Among the 6
patients who were treated for adrenal metastases,
pelvic lymph node metastases, or abdominal lymph
node metastases, 2 patients experienced no discernable toxicity (excluding grade 1 fatigue and skin toxicity); toxicity included grade 1 vaginal bleeding that
resolved (n 5 1 patient), grade 2 diarrhea (n 5 1
patient), grade 2 nausea (n 5 1 patient), and grade 2
flank pain (n 5 1 patient). Among the 15 patients
who were treated for bone metastases, 11 patients
experienced no discernable toxicity (excluding grade
1 fatigue and skin toxicity); toxicity included grade 1
nausea (n 5 1 patient with sternal metastasis), grade
1 cough (n 5 1 patient with sternal metastasis),
grade 2 dysphagia (n 5 1 patient with thoracic spine
metastases), and grade 1 alopecia (n 5 1 patient
with a skull metastasis).
Outcome
One patient with a 397-mL liver metastasis progressed during radiation therapy and died shortly
thereafter, and 35 patients had a documented local
failure from 1 month to 54 months (median, 9
months) after the completion of SBRT. For these 36
patients, the median and mean net GTV was 36 mL
and 80 mL, respectively, versus 26 mL and 40 mL for
patients without a documented local failure
(P 5 .007; 2-tailed t test). Among these 36 patients,
29 also developed distant failure, and 1 was not
assessable for distant failure. In 20 patients, the distant failure was synchronous with local failure, 6
patients developed distant failure from 1 month to
19 months after local failure, and 3 patients had distant failure discovered from 1 month to 6 months
before local failure. In addition to the patient who
progressed during radiation, 2 patients died from
local progression (1 with brain metastases and the
other with liver metastases).
Among the patients who had local failure after
SBRT, 11 patients underwent a second course of curative-intent treatment for their locally recurrent disease. Several patients were retreated with palliativeintent SBRT. Curative-intent salvage therapy included
SBRT (n 5 5 patients), resection (n 5 3 patients),
SRS (n 5 1 patient), liver transplantation (n 5 1
patient), and embolization (n 5 1 patient). Three
patients remained without evidence of active disease
at 29 months, 47 months, and 62 months. In addition
to the 6 patients who underwent attempted curative
reirradiation, 25 patients received additional courses
of SBRT for further oligometastatic disease.
The OS and PFS curves are depicted in Figure 1.
The median survival (MS) was 24 months, and the
median PFS was 11 months. The 2-year and 4-year
OS rates were 50% and 28%, respectively;, and the 2year and 4-year PFS rates were 26% and 20%, respectively. The median PFS, which was calculated so that
salvaged local failures were not considered events,
was 12 months; and the 2-year and 4-year PFS rates
were 31% and 21%, respectively. The LC and distant
control (DC) curves are depicted in Figure 2. The 2year and 4-year patient LC rates were 67% and 60%,
respectively. The actuarial LC rates, which measured
the control of individual lesions, were 77% and 73%
at 2-years and 4-years, respectively. The 2-year and
4-year DC rates were 34% and 25%, respectively.
Fifty-seven patients survived for 2 years, and 36 of
those patients were alive at the last follow-up. Table
2 summarizes these patients’ characteristics. Twenty-
Curative SBRT for Oligometastases/Milano et al.
655
TABLE 2
Characteristics of Long-term (‡2 Years) Survivors
Characteristic
No. of patients (%)
Characteristic
No. of patients (%)
Total no. of patients
No. alive at last follow-up
No. with no evidence of disease
Follow-up of living patients, mo
Range
Median
Deceased, with survival 2 y
Survival, mo
Median [range]
Age, y
Median [range]
Mean SD
Primary cancer
Breast
Colorectal
Lung, head and neck, or esophagus
Pancreas, biliary tract, or liver
Sarcoma
Other
Primary histology
Adenocarcinoma
Squamous cell carcinoma
Sarcoma
Hepatocellular carcinoma
Renal cell carcinoma
Carcinoid
Other
57
36
23
Previously had > 5 metastatic lesions
Prior curative-intent local therapy
Resection
Radiofrequency ablation
Radiation therapy
Cranial stereotactic radiosurgery
Sites involved with oligometastatic disease
Lung
Thoracic lymph nodes
Liver
Pelvic or abdominal lymph nodes
Brain
Adrenal glands
Bone
No. of oligometastatic lesions
1
2
3
4
5
No. of involved organs
1
2
3
Sum of GTVs, mL
Median [range]
Mean SD
8 (14)
20 (35)
17 (30)
2 (4)
1 (2)
0
27–77
41
21
29 [24–55]
57 [35–85]
57 21
23 (40)
15 (26)
7 (12)
1 (2)
4 (7)
y
7 (12)
45 (79)
1 (2)
{
4 (7)
1 (2)
2 (4)
2 (4)
{
2 (4)
23 (40)
11 (19)
25 (44)
2 (4)
2 (4)
0
7 (12)
18 (32)
18 (32)
11 (19)
5 (9)
5 (9)
45 (79)
11 (19)
1 (2)
19 [0.3–403]
38 64
SD indicates standard deviation; GTV, gross tumor volume.
Nine patients remain alive, 6 without evidence of recurrent disease, with a follow-up <2 years (range, 7–17 months); they are not included in this table.
y
Other primary cancers included carcinoid (n 5 2), renal (n 5 2), endometrial, endocervical, and adrenal cortical carcinoma.
{
Sarcoma subtypes included leiomyosarcoma (n 5 3) and Ewing sarcoma. The other histologies were large cell neuroendocrine and cortical carcinoma.
nine patients remained alive and free of detectable
disease for from 7 months to 77 months (median, 36
months).
UVA and MVA
Tables 3 and 4 summarize UVA and MVA for measured outcomes of OS, PFS, DC, and LC. Calculating
PFS so that salvaged local failures were not considered events did not affect which variables were significant. Neither the numbers of organs involved nor
the numbers of oligometastatic lesions were significant for the measured outcomes. Patients who had
primary breast cancer (vs other primary sites) had
significantly improved OS, PFS, LC, and DC. On UVA,
the MS was 20 months (vs not reached), and the 2year and 4-year survival rates were 38% and 18%,
respectively (vs 72% and 64%, respectively) in
patients with breast cancer. On UVA and MVA,
patients with primary pancreatic, biliary or hepatic
cancer (vs other sites) had significantly worse survival and DC. Patients with adrenal metastases
experienced worse OS, PFS, and DC. On UVA, the
MS was 5 versus 24 months, favoring patients without adrenal metastases. On MVA, the net GTV was
significant for all measured outcomes.
Additional Analyses
MVA were rerun with the following time intervals as
continuous variables: from the diagnosis of primary
cancer to the diagnosis of metastatic disease, from
the diagnosis of metastatic disease to the development of oligometastases treated on protocol, and
from the development of oligometastases treated on
protocol to enrollment. These were not significant
for OS or PFS. Age was not significant on MVA for
any measured outcome. On UVA for DC, the occurrence of local failure was not associated with a significantly decreased rate of DC (P 5 .28). However,
on UVA for LC, the presence of distant failure was
associated with worse LC (P 5 .005) and remained
significant on MVA (P 5 .040). In other words,
patients who fail distantly are not more likely to
656
CANCER
February 1, 2008 / Volume 112 / Number 3
TABLE 3
Univariate and Multivariate Analyses for Overall Survival
TABLE 4
Univariate and Multivariate Analyses of Progression-Free Survival,
Local Control, and Distant Control*
P
Variable
UVA
History of >5 metastases prior to enrollment
History of prior curative treatment for oligometastatic disease
No. of organs involved
No. of treated lesions
Primary cancer site*
Colorectal
Lung, head and neck, esophagus (no*)
Pancreas, liver, biliary tract (no*)
Breast (yes*)
Sarcoma
Histology*
Adenocarcinoma (yes*)
Squamous cell carcinoma (no*)
Hepatocellular carcinoma
Renal cell carcinoma
Sarcoma
Carcinoid
Treated sites*
Lung
Thoracic lymph nodes
Liver
Abdominal/pelvic lymph nodes
Adrenal (no*)
Brain
Bone (no*)
Net GTV (smaller*)
.42
.64
.50
.43
P
MVA
UVA
0.32
.007y
.011y
<.0001y
.48
0.017y
.012y
.53
.35
.48
.33
.25
.58
.99
.80
<.00001y
.089
.022y
{
Variable
.75
.035y
.002y
.68
.15
<.0001y
.34
.006y
UVA indicates univariate analysis; MVA, multivariate analysis; GTV, gross tumor volume.
* For variables that were significant on UVA, this characteristic was a favorable predictor for survival.
y
Significant (P < .05).
{
Because this was a continuous variable, it was not assessed in the UVA.
develop or to have developed local failure, whereas
patients who fail locally are at a greater risk of failing
distantly as well.
DISCUSSION
We previously published results on 69 patients with
174 oligometastatic liver lesions who were treated
with SBRT.16 In that series, the 20-month LC rate of
all treated lesions was 57%, and the MS was 14.5
months. We also previously published our results in
49 patients with 125 oligometastatic lung lesions
who were treated with SBRT.17 In those patients, the
3-year LC rate of the treated tumors was 91%, and
the 3-year LC rate of treated patients was 82.5%. In
patients who had 5 lesions confined to the thorax,
the MS was 23.4 months versus 12.4 months in
patients who had extrathoracic disease and/or > 5
sites treated (P < .05). We also separately examined
18 patients who were treated for adrenal metasta-
Primary cancer sitey
Pancreas, liver, biliary tract (noy)
Breast (yesy)
Histologyy
Adenocarcinoma (yes)y
Treated sitesy
Adrenal (noy)
Bone (yesy)
Brain (noy)
Net GTV (smallery)
PFS
LC
MVA
DC
PFS
LC
.024{ .038{ .007{ .050{ .073
.0009{ .027{ .0007{ .014{ .016{
DC
.034{
.006{
NS
NS
.031{
—
—
.99
.003{
.031{
.031{
—§
NS
NS
NS
—§
.004{
NS
NS
—§
.020{
.43
.023{
.005{
—
—
—
<.0001{
.026{
—
—
0.026{
UVA indicates univariate analysis; MVA, multivariate analysis; PFS, progression-free survival; LC,
local control; DC, distant control; NS, nonsignificant (P > .1); GTV, gross tumor volume.
* All variables listed in Table 3 were tested on UVA; Only variables that were significant on UVA are shown.
y
For variables that were significant on UVA, this characteristic was a favorable predictor for survival.
No variable was a significantly adverse predictor for 1 outcome (PFS, LC, or DC) and a significantly
favorable predictor for another.
{
Significant (P > .05).
§
Because this was a continuous variable, it was not assessed in the UVA.
ses.22 The crude LC rate in those patients was 85%,
whereas nearly all patients suffered distant failure.
The MS of the 12 patients in that series who were
treated with curative intent (some of whom were
treated concurrently for liver or lung lesions) was 7
months. The results from our current series suggest
that patients with adrenal metastases fare significantly worse with respect to OS and PFS. However,
because only 2 of 121 patients were treated for adrenal metastases in the current study, this finding must
be interpreted with caution. Nevertheless, the poor
outcome is consistent with our retrospective data.
Patients with brain metastases also fared poorly with
respect to PFS, albeit with small numbers of patients.
There have been several single-institutional series in which patients were treated for oligometastatic lung lesions or oligometastatic liver lesions
with SBRT. These are described in detail elsewhere.11,13 In our prior reports and in others’ series,
toxicity was minimal.
The current report examines all patients who were
enrolled in 2 pilot studies who initially presented with
5 metastatic lesions, regardless of which organs were
involved. Currently, there are very few studies that
address a similar patient population.
Rush University retrospectively studied 23 patients with nonsmall cell lung cancer who were treated aggressively for oligometastatic disease, which
Curative SBRT for Oligometastases/Milano et al.
the authors defined as 1 or 2 metastatic sites.23 Surgery and/or radiation was used. Metastases were
confined to 1 organ, and most patients (n 5 14) had
brain metastases. All but 3 patients had 1 metastatic
site. Their MS was 20 months, and 5 patients
remained alive beyond 3 years.
The University of Chicago is enrolling patients
with from 1 to 5 oligometastases on a Phase I study
with escalating doses from 30 Gy to 42 Gy over 3
fractions. In a preliminary analysis, treated lesions
were controlled well.24 Survival results from that
study are pending.
In our current series, women with oligometastatic disease from primary breast cancer had a significantly better outcome versus other patients. We
previously demonstrated that women with lung metastases from breast cancer fared better with respect
to survival,17 and the current results confirmed this
finding in a population of patients with oligometastatic disease to different sites. It is not known
whether the improved outcome in women with oligometastatic breast cancer reflects a better response
to radiation treatment or whether these findings are
attributable to a better overall prognosis of metastatic breast cancer and more durable disease control
with systemic therapy. We hypothesize that the availability of effective systemic therapy to control subclinical disease, combined with SBRT for clinically
detectable disease, is the basis for this success. A
recently open Southwest Oncology Group (SWOG)
trial will study SBRT in women with oligometastatic
breast cancer: Perhaps a larger cohort will help
answer this question.
It is noteworthy that neither the number of
organs involved nor the number of detectable oligometastatic lesions at enrollment significantly correlated with OS or PFS. Despite the possibility of being
underpowered to detect such differences, the latter
finding is consistent with our previous discovery that
men with prostate cancer who have from 1 to 5 bony
metastases fare equally well regardless of the number
of metastases with which they present.18 In a study
from the University of Pittsburgh in which 205
patients with 4 brain metastases underwent SRS,
net treatment volume was significant for OS and LC,
whereas the number of lesions (range, 4–18 lesions)
was not significant.25 Our current series also suggests
that patients with a history of >5 metastases who
present with 5 apparent metastases at the time of
enrollment do not fare significantly worse in any
measured outcome compared with patients who
present initially with 5 metastases. We previously
postulated that a complete or near-complete response
to systemic therapy potentially may ‘downstage’
657
patients to an oligometastatic state by eradicating
micrometastatic disease.17 Our current analysis suggests that systemic therapy also has the potential to
‘downstage’ patients with numerous clinically apparent metastases to an oligometastatic state.
It is particularly worth noting that, in our study,
the net GTV was significant for OS, PFS, DC, and LC,
consistent with the hypothesis that a larger tumor
burden has a greater risk of local failure as well as
increased metastatic potential. The net GTV in
patients who experienced local failure was significantly greater than the net GTV in patients who did
not experience local failure. Our data suggest that
the risk of developing distant failure is not increased
significantly by the occurrence of local failure, possibly reflecting the magnitude of patients who fail distantly despite LC. Distant failure is a competing risk
with local failure, and, in some patients, the cancer
will fail distantly before it has the opportunity to fail
locally. Our data also suggest that patients who fail
locally are more likely to fail distantly, suggesting the
possibility of a greater risk of distant spread from
local failure. An alternative explanation is that treatment-resistant tumors can synchronously fail locally
and distantly. In the current study, only 3 patients
died from local progression without distant progression, and 3 patients had successfully salvaged local
failure without further disease progression. An interesting question, which we were unable to address in
the current analysis, is whether or not a response to
systemic therapy can improve outcome, not only by
virtue of inherent sensitivity to treatment but also by
virtue of reducing tumor bulk.
In summary, the results from our hypothesisgenerating pilot study support the premise that
aggressive local therapy of limited metastases can
result in prolonged life. Conceivably, systemic therapy has the potential to downstage some patients to
an oligometastatic stage, allowing for a chance for
prolonged life or cure with aggressive local therapy.
Arguably, our data suggest that patients with breast
cancer have the greatest potential to derive benefit,
providing a basis for the SWOG study mentioned
above. Larger tumors may require a greater radiation
dose to achieve more satisfactory LC rates. Different
fractionation schemes, perhaps with a greater dose
per fraction, also may help achieve improved LC,
although risks of late toxicity must be considered
carefully with more hypofractionationed regimens,
particularly with bulkier tumors. Certainly, further
studies are needed to assess the impact of radiation
therapy on patient survival, tumor control, and the
658
CANCER
February 1, 2008 / Volume 112 / Number 3
natural history of oligometastatic disease. We propose that the TNM metastasis staging system be
expanded to incorporate an oligometastatic category,
which would then allow and encourage more systematic testing of the oligometastatic hypothesis.
REFERENCES
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
Hellman S, Weichselbaum RR. Oligometastases. J Clin
Oncol. 1995;13:8–10.
Mehta N, Mauer AM, Hellman S, et al. Analysis of further
disease progression in metastatic non-small cell lung cancer: implications for locoregional treatment. Int J Oncol.
2004;25:1677–1683.
Hellman S. Karnofsky Memorial Lecture. Natural history of
small breast cancers. J Clin Oncol. 1994;12:2229–2234.
Withers HR, Lee SP. Modeling growth kinetics and statistical distribution of oligometastases. Semin Radiat Oncol.
2006;16:111–119.
Hellman S, Weichselbaum RR. Importance of local control
in an era of systemic therapy. Nat Clin Pract Oncol. 2005;
2:60–61.
Tait CR, Waterworth A, Loncaster J, Horgan K, Dodwell D.
The oligometastatic state in breast cancer: hypothesis or
reality. Breast. 2005;14:87–93.
Aboulafia AJ, Levine AM, Schmidt D, Aboulafia D. Surgical
therapy of bone metastases. Semin Oncol. 2007;34:206–214.
Kanner AA, Bokstein F, Blumenthal DT, Ram Z. Surgical
therapies in brain metastasis. Semin Oncol. 2007;34:197–205.
Kuvshinoff B, Fong Y. Surgical therapy of liver metastases.
Semin Oncol. 2007;34:177–185.
Sternberg DI, Sonett JR. Surgical therapy of lung metastases. Semin Oncol. 2007;34:186–196.
Carey-Sampson M, Katz A, Constine LS. Stereotactic body
radiation therapy for extracranial oligometastases: does the
sword have a double edge? Semin Radiat Oncol. 2006;16:67–76.
Timmerman RD, Kavanagh BD, Cho LC, Papiez L, Xing L.
Stereotactic body radiation therapy in multiple organ sites.
J Clin Oncol. 2007;25:947–952.
Kavanagh BD, McGarry RC, Timmerman RD. Extracranial
radiosurgery (stereotactic body radiation therapy) for oligometastases. Semin Radiat Oncol. 2006;16:77–84.
Benedict SH, Galvin J, Hinson W, et al.; SBRT Task Group
101 Members. AAPM report. Stereotactic body radiotherapy. Report of Task Group 101. Available at URL: http://
www.aapm.org/org/structure/default.asp?committee_code5
TG101. Accessed August 15, 2007.
15. Schell MC, Bova FJ, Larson DA, et al.; Task Group 42.
AAPM report no. 54. Stereotactic body radiotherapy. Report
of Task Group 42, 1995. Available at URL: http://www.aapm.
org/pubs/reports/rpt_54.pdf. Accessed August 15, 2007.
16. Katz AW, Carey-Sampson M, Muhs AG, Milano MT, Schell
MC, Okunieff P. Hypofractionated stereotactic body radiation therapy (SBRT) for limited hepatic metastases. Int J
Radiat Oncol Biol Phys. 2007;67:793–798.
17. Okunieff P, Petersen AL, Philip A, et al. Stereotactic body
radiation therapy (SBRT) for lung metastases. Acta Oncol.
2006;45:808–817.
18. Singh D, Yi WS, Brasacchio RA, et al. Is there a favorable
subset of patients with prostate cancer who develop oligometastases? Int J Radiat Oncol Biol Phys. 2004;58:3–10.
19. O’Dell WG, Schell MC, Reynolds D, Okunieff R. Dose
broadening due to target position variability during fractionated breath-held radiation therapy. Med Phys. 2002;29:
1430–1437.
20. Biswas T, Sandhu AP, Singh DP, et al. Low-dose radiosurgery for benign intracranial lesions. Am J Clin Oncol. 2003;
26:325–331.
21. Therasse P, Arbuck SG, Eisenhauer EA, et al. New guidelines to evaluate the response to treatment in solid tumors.
European Organization for Research and Treatment of
Cancer, National Cancer Institute of the United States,
National Cancer Institute of Canada. J Natl Cancer Inst.
2000;92:205–216.
22. Chawla S, Okunieff P, Chen Y, Katz AW, Schell MC, Milano
MT. Stereotactic body radiation therapy (SBRT) for the
treatment of adrenal metastases. Presented at the First
International Symposium on Stereotactic Body Radiation
Therapy and Stereotactic Radiosurgery, Lake Buena Vista,
Fla, February 23–25, 2007.
23. Khan AJ, Mehta PS, Zusag TW, et al. Long term diseasefree survival resulting from combined modality management of patients presenting with oligometastatic, nonsmall cell lung carcinoma (NSCLC). Radiother Oncol.
2006;81:163–167.
24. Weichselbaum RR, Salama JK, Mehta N, et al. Phase I dose
escalation of hypofractionated radiation therapy for
patients with oligometastasis. Presented at the 48th Annual
Meeting of the American Society for Therapeutic Radiation
and Oncology (ASTRO), Philadelphia, Pa, November 5–9,
2006. Abstract 2564.
25. Bhatnagar AK, Flickinger JC, Kondziolka D, Lunsford LD.
Stereotactic radiosurgery for 4 or more intracranial metastases. Int J Radiat Oncol Biol Phys. 2006;64:898–903.